Applied Statistics - Janet Rosenbaum

February 26, 2006

This week, the Applied Statistics Workshop will present a talk by Janet Rosenbaum, a Ph.D. candidate in the Program on Health Policy at Harvard. She majored in physics as an undergraduate at Harvard College and received an AM in statisics last year. Janet will present a talk entitled " Do virginity pledges cause virginity?: Estimating the efficacy of sexual abstinence pledges". She has a publication forthcoming in the American Journal of Public Health on related research. The presentation will be at noon on Wednesday, March 1 in Room N354, CGIS North, 1737 Cambridge St. Lunch will be provided. The abstract of the paper follows on the jump:

Objectives: To determine the efficacy of virginity pledges in delayingsexual debut for sexually inexperienced adolescents in the NationalLongitudinal Study of Adolescent Health (Add Health).
Methods: Subjects were virgin respondents without wave 1 pledge who
reported their attitudes towards sexuality and birth control at wave 1
(n=3443). Nearest-neighbor matching within propensity score calipers
was used to match wave 2 virginity pledgers (n=291) with non-pledgers,
based on wave 1 attitudes, demographics, and religiosity. Treatment
effects due to treatment assignment were calculated.
Results (Preliminary): 17% of virginity pledgers are compliant with their pledge, and do not
recant at wave 3 their earlier report of having taken a pledge. Similar
proportions of virginity pledgers and non-pledgers report having had
pre-marital sex (54% and 61%, p=0.16) and test positive for chlamydia
(2.7% and 2.9%, p=0.89).
Conclusions: Five years after taking a virginity pledge, most virginity
pledgers fail to report having pledged. Virginity pledges do not affect
the incidence of self-reported pre-marital sex or assay-determined
chlamydia.

Posted by Mike Kellermann at February 26, 2006 4:20 PM